A caterpillar's life is not an easy one. The plants that it eats make toxins to make it sick. Birds swoop in to pluck it away and feed it to their chicks. But the most horrific threat comes from wasps that use caterpillars as hosts for their young. These parasitoid wasps are among my favorite creatures (see my post on the emerald cockroach wasp, which attacks cockroaches like a neurosurgeon). So it was with eye-popping delight that I read a new paper in PLOS Biology One about how another species of wasp in Brazil attacks another caterpillar. Glyptapanteles glyptapanteles is more than just cruel to its host. It also gives its host an extreme case of Stockholm syndrome.
The fun begins when a female Glyptapanteles wasp comes across a potential host--a moth known as Thyrinteina leucocerae. The wasp inserts a stinger-like probe into the caterpillar's gut body cavity and injects dozens of eggs. The eggs hatch and grow into wasp larvae, which feed on the still-living host as it continues munching on leaves. The caterpillars even moult and pass through three or four stages with the parasites lurking inside them. Finally, when the wasps have finished their living feast, about 80 of them drill escape holes and crawl out of the caterpillar. They move a few inches away, where they spin cocoons on a twig or leaf, where they will develop into adults.
Many species of wasps exit their hosts this way, and in many cases the hosts promptly die. After you've just spent a couple weeks with dozens of parasites sucking your insides dry and then drilling their way out of your body, you'd probably feel like dying too. But in the case of Thyrinteina, death waits. The caterpillar stops feeding and crawling and simply sits, still alive, next to the wasp cocoons. When other insects come by, they wave their heads violently around, so violently they can knock the other insects off the tree. (You can download a movie of this behavior here.) Once the adult wasps emerge from the cocoons, the caterpillars finally expire.
You may be thinking that the caterpillars were turned into bodyguards for their parasites, protecting them from predators. And if it were ture, it would be a particularly striking addition to a long list of cases in which parasites manipulate their hosts for their own well being. I wrote about this manipulation at length in my book Parasite Rex, and have also written about it here on the Loom (suicidal rats, neurosurgical wasps, etc).
When scientists find a host acting weirdly, it's a reasonable hypothesis that they're being manipulated by their host parasite. But it's just a hypothesis, one that cries out for testing. There are other possible interpretations, after all. W,hat looks like a clever adaptation that boosts the parasite's reproductive success may in fact just by an incidental byproduct of being sick. And a peculiar behavior that scientists observe in hosts they keep in a lab may not be terribly important out in the wild. Parasites sometimes need to go from one host to another to develop--in many cases, traveling from prey to the predators that eat them. If a parasite makes its host an easier target for predators, its host may get eaten by the wrong species of predator, one in which the parasite will die.
To test the bodyguard hypothesis on Glyptapanteles wasps, scientists ran experiments on the animals outdoors, on real guava trees. They observed how parasitized caterpillars behaved around the cocoons compared to healthy ones, and they observed how much protection the infected caterpillars really provided from predators by removing some of them.
The scientists found that almost all the parasitized caterpillars huddled close to their parasite cocoons, often actually crouching over them. When the scientists put cocoons next to unparasitized caterpillars, they simply went on wandering the leaves and twigs, feeding away. When stinkbugs came along, the parasitized caterpillars almost always lashed out, while the healthy caterpillars ignored them. Sometimes the stinkbugs and wasps fell off the tree, and sometimes they just gave up. This flailing made a big difference to the survival of the wasps, the scientists found. Removing infected caterpillars from the neighborhood of the cocoons double the death rate for the wasps.
So the bodyguard hypothesis looks good. But it naturally raises a question: how do the wasp larvae turn the caterpillars into their bodyguards? The catepillars only start fending off predators a couple hours after the wasps had left their bodies. It's possible that they left behind some chemicals that altered their hosts' behavior. But when the scientists dissected caterpillars three or four days after the wasps had left, they would find one or two living wasp larvae still inside. Scientists have found other parasites that have stayed behind in their hosts. Ants, for example, are infected by lancet flukes that drive them up to the tops of blades of grass, where they can be eaten by grazing livestock. It appears that they are driven by a few flukes that form cysts in the brains of the ants and can not pass on into their new host like the rest of the flukes. Perhaps the bodyguard caterpillars are being piloted by one or two wasps that stay behind, defending their siblings from predators while surrendering their own lives. Who knew such vicious parasites could be so heroic?
(Photograph by Prof. José Lino-Neto. Covered by a Creative Commons Attribution License. Any use should include citation of the authors and paper as the original source.)
[Note: Thanks for the fact-checking and copy-editing. I've fixed the text accordingly.]
Life Science
Brain & Behavior
Comments
Just a little note - it's PLoS ONE (not PLoS Biology) and you can send a trackback if you want.
Posted by: Coturnix | June 3, 2008 8:21 PM
Kinda like the fraction of cells in a slime mold which give up their chance at making spores. . . I know some evolution-of-altruism people who would eat this example up. :-)
Posted by: Blake Stacey | June 4, 2008 10:23 AM
This is very cool. It is surprising how many parasites seem to use the strategy of sacrificing a few to get control of a host. Toxoplasma does this also. This also could evolve from inefficiency: if a parasite was leaving a few copies behind in a host or a few were going to a different location then the parasite might evolve a way of taking advantage of that inefficiency. By the time we see it then the entire system looks like an adaptation.
Incidentally, you may want to fix "ture" "W,hat" and "cateripillars."
Posted by: Joshua Zelinsky | June 4, 2008 11:52 AM
I liked this comment, but it contains some errors.
For example, the parasitoid does not oviposit into the caterpillar's gut, but into its heamolymph.
Also, we the scientists did not send in stinkbugs outside, we simply exposed the wasp pupae to the naturally occuring predators and parasitoids. We did this with groups of pupae from which we removed the caterpillar and groups with the bodyguard caterpillar.
See the original publication at PLoS ONE for more details (http://www.plosone.org/doi/pone.0002276 will take you directly to the publication)
Posted by: Arne Janssen | June 4, 2008 4:41 PM
Very cool article -- found an error, though. "When scientists find a host acting weirdly, it's a reasonable hypothesis that they're being manipulated by their host." The second "host" should be... eh... "parasite," I suppose.
Posted by: iratecat | June 4, 2008 10:00 PM
The text mentions "about 80" of the parasites crawl out of the caterpillar. The image shows far less eggs near the host however, and I cannot imagine 80 of them fitting in that one host. Are you sure 80 is the right number?
(P.S. The typos mentioned by Joshua Z. above are still not fixed)
Posted by: Kilian Hekhuis | June 5, 2008 9:50 AM
Video link no longer works. Found one here: http://www.youtube.com/watch?v=zN37VfOPbxE
Posted by: Daro | June 5, 2008 10:27 AM
Very interesting. I am more of an ants fan, but these insects are fascinating, too. I have a couple of questions regarding this behavior though, if you do not mind.
1. Could it be that the moth thought that the parasites were its own eggs or something? I remember seeing how a certain predator of ants could somehow pretend to be one of their own ants using its antennas to mimic the ants'. Walked straight into a nest full of ants like that and managed to survive without as much as a scratch. That is so cool.
2. How did the scientists rule out the "byproduct of being sick" theory? Comparing the behavior of the "infected" vs "healthy" does not rule that out, does it? I mean, if they were not "infected" then they will not shown any of these signs in the first place, right?
3. I skimmed through the paper and while it did mention several times that the scientists removed the infected caterpillars, did they also verify that it was not showing any "bodyguard" behavior after they were removed? If no such behavior was detected after they were removed, would that also not be a significant argument to support the bodyguard theory? In that case, have I somehow missed it in the paper? If they still showed "bodyguard" behavior after they were removed, would that not support the "byproduct of being sick" rather than the "bodyguard" theory?
Posted by: Wobbler | June 6, 2008 11:52 AM
Ah, never mind. Just read this from the paper.
'The host remains alive, stops feeding and moving, spins silk over the pupae, and responds to disturbance with violent head-swings (supporting information).'
I guess the host spinning silk over the pupae is pretty conclusive concerning the "bodyguard" theory. Although I am still wondering whether the violent head-swings is not just a natural defensive change after having its inside eaten?
Posted by: Wobbler | June 6, 2008 4:58 PM
When I read the article I tough "this must be on The Loom". Anyway I have posted about it to and uploaded the videos of infected and uninfected caterpillar on Youtube:
http://www.youtube.com/watch?v=6lk_m1AF4Xk&feature;=related
http://www.youtube.com/watch?v=RAeYPCQoqy4&feature;=related
Posted by: Atila | June 17, 2008 3:39 PM
"W,hat looks like a clever adaptation that boosts the parasite's reproductive success may in fact just by an incidental byproduct of being sick."
What's the difference? There isn't any decision making that goes into causing this behavior, it wasn't *chosen* as an adaptation, so it's going to be a by-product.
Is the question whether the actions are beneficial or not?
Posted by: Ian Calvert | June 18, 2008 12:03 PM
Awesome read.
Posted by: website design | June 18, 2008 12:16 PM
Ian Calvert: I interpret that as asking whether the behaviour is directly caused simply by having been infected (the protection of the wasps being incidental), or if the wasps alter the host in some further way to cause the behaviour.
Posted by: Kristoffer | June 18, 2008 1:50 PM
Has it been observed that the caterpillar's dead body is being manipulated by the wasps larvae or is it assumed since there are some larvae they must be doing it.
Ps this is not obvious to me am not a bio student.
Posted by: wowmir | June 18, 2008 1:54 PM
Ian: Insightful. The question is of a deeper understanding of the mechanisms involved. One wants to know whether this is the natural behavior of the caterpillar that just happens to be beneficial for wasps or something that's specifically caused be the wasp larva. If its something caused by the wasp larva, then a study of how the wasp is controlling the caterpillar could prove beneficial in other areas.
Posted by: Bob/Paul | June 18, 2008 1:58 PM
Well, Carl, you Darwinist roader, what evolutionary pathway do you suggest for this particular case of parasitism? ("Darwinist roader" comes from the old Communist Chinese expression, "capitalist roaders and their running dogs").
Your problem, Carl, is that you are aware of complex interspecies relationships such as this, yet you are not skeptical of evolution theory. You even wrote a book titled "Evolution: The Triumph of an Idea."
I wrote on my blog,
Previously my arguments about co-evolution were restricted to the evolution of obligate mutualism (i.e., total co-dependence of two different kinds of organisms) because I thought that only co-evolution of obligate mutualism could require that a mutation in one kind of organism be immediately answered by a corresponding mutation in another kind of organism at the same geographical location in order to produce a benefit or even just for survival. I didn't see parasitism per se as a problem for evolution because I assumed that mutations involving parasitism do not require an immediate corresponding mutation in the other organism. However, I have just discovered literature about bizarre parasitisms that may require changes in the traits of the host and so also may be a problem for evolution.
My argument about co-evolution was originally stated as follows:
"In the co-evolution of obligate mutualism, unlike in evolutionary adaptation to widespread fixed physical features of the environment, e.g., air, water in its different forms, and land its many forms, there may be nothing to adapt to, and the reason why there may be nothing to adapt to is that the corresponding co-dependent trait in the other organism is likely to be initially absent locally."
I present "buzz" pollination as a good example of this dilemma of the co-evolution of obligate mutualism.
Anyway, thanks very much, Carl, for bringing this example of parasitism to my attention. I have added it to my group of posts under the post label "Non-ID criticisms of evolution" on my blog.
Posted by: Larry Fafarman | June 18, 2008 2:03 PM
Thank you for the thought provoking replies.
Kristoffer: I see, although I don't really agree with the word "incidental". A distinction between there being one or two stages would be interesting.
Bob/Paul: I gathered from the summary that this was not healthy behavior, but I think that it may be a distinction between a general behavior when sick and a behavior when sick *in this particular manner*. Very interesting all around :)
I'll have to add the paper to my reading list!
Posted by: Ian Calvert | June 18, 2008 10:31 PM
Wow, now that was some fascinating stuff. No doubt about it.
www.FireMe.To/udi
Posted by: John Boy Toomey | June 19, 2008 12:02 AM
Typos: "W,hat" and "ture"
Posted by: Carl | June 19, 2008 4:55 AM
Just a little note "And if it were ture" should be "And if it were true."
Posted by: Fooster | June 19, 2008 7:08 AM